European Statistics PhD

[miserablefunction]

I am thinking of applying statistics PhD programs in Europe(UK and Switzerland) and would appreciate a lot if anyone could answer the following questions. 1) How is statisics PhD perceived in US or Canada? I know ETH Zurich phds are perceived very well in US. 2) With master’s curriculum in north america, would there be any problem following their PhD program? 3) I can’t really find the number of admitted students each year, anyone knows their application details? 4) I also find that many programs specify their deadline as like XX weeks/months prior to your intended start date, does this mean they admit PhD students in rolling base? 5) Is the funding in UK similar to that of US/Canada for international students? 

[Stat Assistant Professor]

Assuming you have publications (either in press or submitted), a PhD from any of the well-regarded schools in Europe (Oxford, Cambridge, ETH Zurich, Warwick, etc.) would make you competitive in the academic job market in the U.S. and Canada. I'm not as certain about industry, but securing a job in academia in the U.S. or Canada with a PhD from a good European would typically not be a problem. Additionally, the reputation of the PhD advisor and publications are the most important thing. Someone who works with Peter Buhlmann or Sara van de Geer, for example, and publishes journal articles as a PhD student should have almost no difficulty landing a good postdoc in the U.S.

I don't know much about admissions in Europe, but the PhD programs there contain no coursework, only research. So you would spend 3-4 years just doing research and would need to teach yourself all the material needed for your research (but every PhD student in the U.S. has to more-or-less teach themselves their research area/topic, and everyone has some guidance from their PhD supervisor in the beginning). I do get the impression that Statistics in Europe is quite theoretical in general, though, so they may prefer to admit students who have already taken classes like measure theoretic probability, advanced statistical inference, and more advanced theoretical courses (classes that would typically be part of the PhD curriculum in the U.S., but not part of the Masters curriculum here).

Edited August 29, 2018 by Applied Math to Stat

[miserablefunction]

On 8/29/2018 at 9:23 PM, Applied Math to Stat said:

Assuming you have publications (either in press or submitted), a PhD from any of the well-regarded schools in Europe (Oxford, Cambridge, ETH Zurich, Warwick, etc.) would make you competitive in the academic job market in the U.S. and Canada. I'm not as certain about industry, but securing a job in academia in the U.S. or Canada with a PhD from a good European would typically not be a problem. Additionally, the reputation of the PhD advisor and publications are the most important thing. Someone who works with Peter Buhlmann or Sara van de Geer, for example, and publishes journal articles as a PhD student should have almost no difficulty landing a good postdoc in the U.S.

I don't know much about admissions in Europe, but the PhD programs there contain no coursework, only research. So you would spend 3-4 years just doing research and would need to teach yourself all the material needed for your research (but every PhD student in the U.S. has to more-or-less teach themselves their research area/topic, and everyone has some guidance from their PhD supervisor in the beginning). I do get the impression that Statistics in Europe is quite theoretical in general, though, so they may prefer to admit students who have already taken classes like measure theoretic probability, advanced statistical inference, and more advanced theoretical courses (classes that would typically be part of the PhD curriculum in the U.S., but not part of the Masters curriculum here).

How do you compare LSE, Imperial and UCL with Warwick? Are they comparable?

[Stat Assistant Professor]

4 hours ago, miserablefunction said:

How do you compare LSE, Imperial and UCL with Warwick? Are they comparable?

Those are all very good schools overall, but I am not sure how their Statistics departments compare to each other. TBH, when it comes to the UK and Europe, I am more familiar with individual "big shot" names (e.g. van de Geer, van der Vaart, etc.) than I am with the departments themselves. It's often been said here that the PhD advisor's reputation matters more than the institution. I would have to agree. Someone with a renowned professor as their PhD advisor and one or two papers in respectable journals has a very good chance at securing a top postdoc in the U.S. Even if the advisor is not as well-known, having one paper published, accepted, or in revision at JRSS-B, JASA, AoS, or Biometrika gives you a huge advantage in the academic job market. My department extended a TT job offer to someone this past year mainly because they were the single author for an Annals paper (this person ultimately accepted a TT job offer from Purdue) -- this person was from Michigan State, if I recall correctly. TL;DR: the quality of publications and the recommendation letters from PhD advisor(s) and postdoc mentors seem to matter the most for academic employment. PhD granting institution is secondary to these two things.

Edited September 1, 2018 by Applied Math to Stat